Improved Physics-Informed Neural Network based AC Power Flow for Distribution Networks

被引:2
作者
Eeckhout, Victor [1 ]
Fani, Hossein
Hashmi, Md Umar
Deconinck, Geert
机构
[1] Katholieke Univ Leuven, Genk, Belgium
来源
2024 IEEE PES INNOVATIVE SMART GRID TECHNOLOGIES EUROPE, ISGT EUROPE | 2024年
关键词
AC power flow; physics-informed neural network; model-free techniques; data-driven model; distribution networks;
D O I
10.1109/ISGTEUROPE62998.2024.10863674
中图分类号
TP39 [计算机的应用];
学科分类号
081203 ; 0835 ;
摘要
Power flow analysis plays a critical role in the control and operation of power systems. The high computational burden of traditional solution methods led to a shift towards data-driven approaches, exploiting the availability of digital metering data. However, data-driven approaches, such as deep learning, have not yet won the trust of operators as they are agnostic to the underlying physical model and have poor performances in regimes with limited observability. To address these challenges, this paper proposes a new, physics-informed model. More specifically, a novel physics-informed loss function is developed that can be used to train (deep) neural networks aimed at power flow simulation. The loss function is not only based on the theoretical AC power flow equations that govern the problem but also incorporates real physical line losses, resulting in higher loss accuracy and increased learning potential. The proposed model is used to train a Graph Neural Network (GNN) and is evaluated on a small 3-bus test case both against another physics-informed GNN that does not incorporate physical losses and against a model-free technique. The validation results show that the proposed model outperforms the conventional physics-informed network on all used performance metrics. Even more interesting is that the model shows strong prediction capabilities when tested on scenarios outside the training sample set, something that is a substantial deficiency of model-free techniques.
引用
收藏
页数:6
相关论文
共 24 条
[1]   Probabilistic load flow in distribution networks: An updated and comprehensive review with a new classification proposal [J].
Abbasi, Ali Reza ;
Mohammadi, Mohammad .
ELECTRIC POWER SYSTEMS RESEARCH, 2023, 222
[2]   XGBoost: A Scalable Tree Boosting System [J].
Chen, Tianqi ;
Guestrin, Carlos .
KDD'16: PROCEEDINGS OF THE 22ND ACM SIGKDD INTERNATIONAL CONFERENCE ON KNOWLEDGE DISCOVERY AND DATA MINING, 2016, :785-794
[3]   Neural networks for power flow: Graph neural solver [J].
Donon, Balthazar ;
Clement, Remy ;
Donnot, Benjamin ;
Marot, Antoine ;
Guyon, Isabelle ;
Schoenauer, Marc .
ELECTRIC POWER SYSTEMS RESEARCH, 2020, 189 (189)
[4]   Can Locational Disparity of Prosumer Energy Optimization Due to Inverter Rules Be Limited? [J].
Hashmi M.U. ;
Deka D. ;
Busic A. ;
Van Hertem D. .
IEEE Transactions on Power Systems, 2023, 38 (06) :5726-5739
[5]  
Hashmi MU, 2024, Arxiv, DOI arXiv:2409.01284
[6]   Robust flexibility needs assessment with bid matching framework for distribution network operators [J].
Hashmi, Md Umar ;
Koirala, Arpan ;
Ergun, Hakan ;
Van Hertem, Dirk .
SUSTAINABLE ENERGY GRIDS & NETWORKS, 2023, 34
[7]   Physics-Guided Deep Neural Networks for Power Flow Analysis [J].
Hu, Xinyue ;
Hu, Haoji ;
Verma, Saurabh ;
Zhang, Zhi-Li .
IEEE TRANSACTIONS ON POWER SYSTEMS, 2021, 36 (03) :2082-2092
[8]   Physics-informed machine learning with optimization-based guarantees: Applications to AC power flow [J].
Jalving, Jordan ;
Eydenberg, Michael ;
Blakely, Logan ;
Castillo, Anya ;
Kilwein, Zachary ;
Skolfield, J. Kyle ;
Boukouvala, Fani ;
Laird, Carl .
INTERNATIONAL JOURNAL OF ELECTRICAL POWER & ENERGY SYSTEMS, 2024, 157
[9]   Decoupled probabilistic feeder hosting capacity calculations using general polynomial chaos [J].
Koirala, Arpan ;
Hashmi, Md Umar ;
D'hulst, Reinhilde ;
Van Hertem, Dirk .
ELECTRIC POWER SYSTEMS RESEARCH, 2022, 211
[10]   Data-Driven Optimal Power Flow: A Physics-Informed Machine Learning Approach [J].
Lei, Xingyu ;
Yang, Zhifang ;
Yu, Juan ;
Zhao, Junbo ;
Gao, Qian ;
Yu, Hongxin .
IEEE TRANSACTIONS ON POWER SYSTEMS, 2021, 36 (01) :346-354